/*
* Handle caching attributes in page tables (PAT)
*
* Authors: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
* Suresh B Siddha <suresh.b.siddha@intel.com>
*
* Loosely based on earlier PAT patchset from Eric Biederman and Andi Kleen.
*/
#include <linux/seq_file.h>
#include <linux/bootmem.h>
#include <linux/debugfs.h>
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/gfp.h>
#include <linux/mm.h>
#include <linux/fs.h>
#include <linux/rbtree.h>
#include <asm/cacheflush.h>
#include <asm/processor.h>
#include <asm/tlbflush.h>
#include <asm/x86_init.h>
#include <asm/pgtable.h>
#include <asm/fcntl.h>
#include <asm/e820.h>
#include <asm/mtrr.h>
#include <asm/page.h>
#include <asm/msr.h>
#include <asm/pat.h>
#include <asm/io.h>
#ifdef CONFIG_X86_PAT
int __read_mostly pat_enabled = 1;
static inline void pat_disable(const char *reason)
{
pat_enabled = 0;
printk(KERN_INFO "%s\n", reason);
}
static int __init nopat(char *str)
{
pat_disable("PAT support disabled.");
return 0;
}
early_param("nopat", nopat);
#else
static inline void pat_disable(const char *reason)
{
(void)reason;
}
#endif
static int debug_enable;
static int __init pat_debug_setup(char *str)
{
debug_enable = 1;
return 0;
}
__setup("debugpat", pat_debug_setup);
#define dprintk(fmt, arg...) \
do { if (debug_enable) printk(KERN_INFO fmt, ##arg); } while (0)
static u64 __read_mostly boot_pat_state;
enum {
PAT_UC = 0, /* uncached */
PAT_WC = 1, /* Write combining */
PAT_WT = 4, /* Write Through */
PAT_WP = 5, /* Write Protected */
PAT_WB = 6, /* Write Back (default) */
PAT_UC_MINUS = 7, /* UC, but can be overriden by MTRR */
};
#define PAT(x, y) ((u64)PAT_ ## y << ((x)*8))
void pat_init(void)
{
u64 pat;
bool boot_cpu = !boot_pat_state;
if (!pat_enabled)
return;
if (!cpu_has_pat) {
if (!boot_pat_state) {
pat_disable("PAT not supported by CPU.");
return;
} else {
/*
* If this happens we are on a secondary CPU, but
* switched to PAT on the boot CPU. We have no way to
* undo PAT.
*/
printk(KERN_ERR "PAT enabled, "
"but not supported by secondary CPU\n");
BUG();
}
}
/* Set PWT to Write-Combining. All other bits stay the same */
/*
* PTE encoding used in Linux:
* PAT
* |PCD
* ||PWT
* |||
* 000 WB _PAGE_CACHE_WB
* 001 WC _PAGE_CACHE_WC
* 010 UC- _PAGE_CACHE_UC_MINUS
* 011 UC _PAGE_CACHE_UC
* PAT bit unused
*/
pat = PAT(0, WB) | PAT(1, WC) | PAT(2, UC_MINUS) | PAT(3, UC) |
PAT(4, WB) | PAT(5, WC) | PAT(6, UC_MINUS) | PAT(7, UC);
/* Boot CPU check */
if (!boot_pat_state)
rdmsrl(MSR_IA32_CR_PAT, boot_pat_state);
wrmsrl(MSR_IA32_CR_PAT, pat);
if (boot_cpu)
printk(KERN_INFO "x86 PAT enabled: cpu %d, old 0x%Lx, new 0x%Lx\n",
smp_processor_id(), boot_pat_state, pat);
}
#undef PAT
static char *cattr_name(unsigned long flags)
{
switch (flags & _PAGE_CACHE_MASK) {
case _PAGE_CACHE_UC: return "uncached";
case _PAGE_CACHE_UC_MINUS: return "uncached-minus";
case _PAGE_CACHE_WB: return "write-back";
case _PAGE_CACHE_WC: return "write-combining";
default: return "broken";
}
}
/*
* The global memtype list keeps track of memory type for specific
* physical memory areas. Conflicting memory types in different
* mappings can cause CPU cache corruption. To avoid this we keep track.
*
* The list is sorted based on starting address and can contain multiple
* entries for each address (this allows reference counting for overlapping
* areas). All the aliases have the same cache attributes of course.
* Zero attributes are represented as holes.
*
* The data structure is a list that is also organized as an rbtree
* sorted on the start address of memtype range.
*
* memtype_lock protects both the linear list and rbtree.
*/
struct memtype {
u64 start;
u64 end;
unsigned long type;
struct list_head nd;
struct rb_node rb;
};
static struct rb_root memtype_rbroot = RB_ROOT;
static LIST_HEAD(memtype_list);
static DEFINE_SPINLOCK(memtype_lock); /* protects memtype list */
static struct memtype *memtype_rb_search(struct rb_root *root, u64 start)
{
struct rb_node *node = root->rb_node;
struct memtype *last_lower = NULL;
while (node) {
struct memtype *data = container_of(node, struct memtype,